Google Chrome now has Dark Mode thanks to Windows 11’s Bing AI, however there are still some limitations

Microsoft has made significant strides in expanding the availability of Bing.com with Bing AI to a wider audience. Initially exclusive to Microsoft Edge users, Bing AI, powered by the advanced ChatGPT-4 language model, is now being rolled out to other browsers, including Google Chrome and Apple’s Safari.

Windows 11’s Bing AI Expands

Users have started receiving pop-ups on their Windows 10 or 11 taskbars, offering them the opportunity to try Bing AI in Chrome. Notably, this version of Bing AI also comes with native dark mode support, similar to the one present in Microsoft Edge. The dark mode provides a sleek black background that matches the aesthetic of Edge’s dark theme.

Within Bing.com, users can easily switch between light and dark themes through the hamburger menu, and there is even a system default option that automatically aligns with the operating system’s preference.

However, it’s worth mentioning that Microsoft has imposed some limitations on Bing Chat capabilities in Chrome. Unlike Edge, which allows up to 30 messages per conversation and 4,000 characters per message, Chrome’s Bing AI is currently limited to five messages per conversation and 2,000 characters per message. Additionally, users may notice prompts encouraging them to explore Bing AI in Edge when accessing it through Google Chrome.

Microsoft is continuously working on improving Bing AI’s capabilities. One notable upcoming feature is “No search” or “ChatGPT-like mode,” which aims to enhance user interaction with Bing AI. Currently, Bing AI relies on traditional web searches to answer user queries, which can sometimes lead to less human-like responses. However, with the “No search” mode, Bing AI will rely more on its AI capabilities, similar to ChatGPT, and provide responses directly from its knowledge base rather than crawling the web for answers.

Recently, Microsoft updated Bing AI with visual search, incorporating OpenAI’s image recognition technology. This enables Bing AI to interpret and describe images, providing relevant answers and information related to the visual content.

It’s essential to note that Bing AI’s support for all browsers is still being gradually rolled out and is currently accessible to select users. However, according to a Microsoft source, Bing AI will become available in Chrome and Safari for everyone in the coming days or weeks.

Despite these advancements, Microsoft ensures that Bing will continue to be free for users, while the enterprise edition of Bing will operate independently, catering to specific business needs.

In conclusion, Microsoft’s efforts to expand Bing AI’s availability and introduce new features like “No search” and visual search show its commitment to enhancing user experience and providing a powerful and efficient AI-driven search engine.

Tesla gives up on selling cars

Google’s AI enabled me to create subpar Nintendo knockoffs

Chevrolet’s Silverado ZR2 looks great, but smart money buys a different trim

OpenAI disputes culpability in the teen suicide action, citing’misuse’ of ChatGPT

Social Tech

Instagram Stories now offers users direct access to Meta AI editing tools

by George Mensah 2025-10-29

Meta is pushing deeper into AI territory with new AI-editing tools in Instagram Stories, where users can edit images and videos simply by typing in what they want to modify. From hair color to special effects, the feature upends the possibilities of creators and regular users alike to personalize their content.

Text Prompts Meet Visual Creativity

Until now, Instagram’s AI editing tools were primarily accessible through Meta AI’s chatbot, which required users to interact via direct messages. With this latest integration, however, AI editing becomes native to Stories, allowing anyone to make instant visual edits using plain language commands.

These new edit features come under the “Restyle” menu that can be accessed using the paintbrush icon in Instagram Stories. One can type commands such as “give me a sunset background,” “remove the person in the corner,” or “color my hair pink.” The AI carries out the edit one wants within seconds.

Meta suggests that users only have three primary actions to select from — Add, Remove, or Change — while specifying what they’d like to alter. The AI will automatically add objects, alter appearances, or completely restyle the photo based on what they’ve described.

Preset Effects and Dynamic Video Edits

In addition to custom prompts, Instagram also has pre-select AI effects that can beautify or stylize posts. Filters like sunglasses, a denim jacket, or even a watercolor art effect can be applied.

On video content, the feature does even better — creators are able to superimpose atmospheric effects like falling snow, glowing embers, or cinematic lighting, which makes Stories appear more polished and professional without the necessity of using editing apps.

Privacy and AI Usage Terms

While the new features enable creativity, they come with privacy implications. Being used to introduce users to Meta’s Terms of Service for AI, which allow the company to “analyze photos and videos, including facial data, to make AI better.” According to Meta, it allows its systems to “summarize image contents, edit images, and generate new content based on the image.”

Critics have also had concerns regarding the ways in which such data might be used to train Meta’s broader AI models, though the company has sworn to remain committed to responsible innovation and transparency.

Meta’s Expanding AI Push

The release of AI editing software is just part of Meta’s overall strategy to roll out artificial intelligence on every platform it has, from Facebook and Instagram to WhatsApp. Recently, Meta began beta-testing a “Write with Meta AI” feature, which helps users compose intelligent or engaging comments under Instagram posts.

Meanwhile, Meta’s separate Meta AI app — with its chatbot and new “Vibes” AI-generated video stream — has been picking up steam. According to Similarweb estimates, iOS and Android daily active users rose from 775,000 to 2.7 million over a four-week span as of October 17.

Protecting Younger Users

As a response to increasing complaints from regulators and parents, Meta has also added new parental tools for its AI features. Parents may now shut off chats with AI characters and filter topics that their teens have with the chatbot to provide a safer online environment.

With these new instruments, Instagram is not only emerging as a social network but a creative platform fueled by generative AI. With Meta, OpenAI, and Google competing for leadership, this launch shows how AI is becoming more a part of the social fabric of our era — blurring the line between creativity, technology, and self-expression.

Social Tech

A program is launched by Meta to promote the adoption of their Llama AI models by companies

by George Mensah 2025-05-22

Meta is rolling out red carpet treatment for AI startups with its new Llama for Startups initiative—offering cash, technical support, and exclusive access to its AI engineering team. But beneath the generous facade lies a fierce battle for dominance in the trillion-dollar generative AI market.

What Startups Get From Meta’s Program

💰 **Up to 36,000∗∗(36,000∗∗(6K/month for 6 months) in cloud credits
🤝 Direct engineering support from Meta’s Llama team
🔧 Early access to custom Llama model fine-tuning tools
🌐 Networking with other AI-first startups

Eligibility requirements are surprisingly accessible:

U.S.-based incorporation
Less than $10M in total funding
At least one developer on payroll
Building generative AI products

Deadline to apply: May 30, 2024

Why Meta Needs Startups More Than Ever

Despite 1 billion+ Llama downloads, Meta faces mounting pressure:

🔥 Competitive Threats

Google’s Gemini and Anthropic’s Claude dominate enterprise adoption
OpenAI’s GPT-4o leads in multimodal capabilities
Mistral, DeepSeek, and Alibaba’s Qwen are winning open-source favor

🚨 Recent Llama Stumbles

Llama 4 Behemoth delayed due to underperformance (WSJ)
Benchmark cheating allegations on LM Arena leaderboard
Public vs. “optimized” model discrepancies eroding trust

💸 Meta’s Make-or-Break AI Bet

Projecting 2B−2B−3B AI revenue in 2025
Banking on 460B−460B−1.4T by 2035 (yes, trillion)
Spending $900M+ annually just on GenAI R&D

The Hidden Strategy Behind the Startup Play

This isn’t just altruism—it’s a three-pronged chess move:

Lock-In Future Customers
Startups that build on Llama today become enterprise buyers tomorrow.
Crowdsource Innovation
Early adopters essentially beta-test new Llama capabilities for free.
Combat Open-Source Defections
With alternatives like Mistral gaining traction, Meta needs to make Llama indispensable.

What’s Really at Stake?

Meta’s playing a long infrastructure game:

60B−60B−80B earmarked for 2025 data centers
Revenue-sharing deals with cloud providers hosting Llama
Future Llama API monetization (Zuck hinted at ads/subscriptions)

For startups, the calculus is simple:
✅ Free money and support in a cash-strapped AI winter
❌ Risk of vendor lock-in as Llama evolves

Should Your Startup Apply?

The case for jumping in:

If you’re already using Llama, this is free acceleration
Early access could provide competitive edge
Meta’s engineering insights are gold dust for product refinement

Reasons to hesitate:

$36K doesn’t go far with today’s GPU costs
Potential IP concerns working closely with a tech giant
Llama’s long-term roadmap remains uncertain

The Bottom Line

Meta’s throwing a Hail Mary to cement Llama as the open-weight model of choice. For scrappy AI startups, it’s a rare chance to piggyback on Meta’s war chest—just don’t mistake it for a long-term partnership.

Load More Posts